A ceramic package basic surface defect detection method and system based on image recognition

Abstract

The application provides a ceramic packaging basic surface defect detection method and system based on image recognition, which performs first visual collection and recognition on the surface of the contact area between an electronic element and a ceramic packaging layer, obtains contact space state data between the two, performs global quantitative recognition on the contact tightness between the two, thereby determining the contact abnormal area between the two, performs image picture extraction and recognition on the contact abnormal area, more finely analyzes the contact interface topography data of the two, characterizes the contact change of the two, thereby determining the contact defect area between the two, realizing positioning of the contact loose area of the electronic element and the ceramic packaging layer, performing second visual collection and recognition on the outer surface of the ceramic packaging layer, obtaining the outer surface defect area of the ceramic packaging layer, and based on the relative position relationship information of the contact defect area and the outer surface defect area, determining the structure defect area of the ceramic packaging layer, thereby globally positioning and recognizing the structure defect of the ceramic packaging layer of the electronic element, and improving the good product discrimination accuracy of the electronic element.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of electronic component manufacturing, and particularly to a ceramic packaging basic surface defect detection method and system based on image recognition. BACKGROUND

[0002] Electronic components such as chips still need to be packaged with ceramic materials after being completed, which can isolate electronic components from the external environment. Ceramic packaging not only isolates electronic components from external air and water vapor, but also isolates electronic components from external electromagnetic interference, thereby improving the heat resistance, corrosion resistance, corrosion resistance and electromagnetic interference resistance of electronic components. The existing ceramic packaging method for electronic components is to use ceramic materials such as aluminum oxide, aluminum nitride or silicon carbide to make a ceramic substrate that matches the size of the electronic component, and then put the ceramic substrate and the electronic component into a high-temperature furnace for sintering, so that the two are tightly combined to form a whole. However, due to the influence of the sintering operation process, it cannot be guaranteed that the ceramic substrate is tightly combined with the electronic component in the entire area, and the ceramic substrate may also form cracks on its surface due to improper sintering operation, which will affect the sealing of the ceramic substrate to the electronic component and cannot form a good protection to the electronic component. Therefore, it is necessary to conduct global non-contact defect detection on the surface of the ceramic packaging of the electronic component, so as to accurately and comprehensively identify the effect of the ceramic packaging, and improve the reliability of the good product identification of the electronic component. SUMMARY

[0003] The purpose of the present application is to provide a ceramic packaging basic surface defect detection method and system based on image recognition, which performs first visual acquisition and identification on the surface of the contact area between the electronic component and the ceramic packaging layer, obtains the contact space state data between the two, globally quantitatively identifies the contact tightness between the two, and determines the contact abnormal area between the two; then image frame extraction and identification are performed on the contact abnormal area to more finely analyze the contact interface topography data of the two, and the contact change of the two is characterized to determine the contact defect area between the two, so as to position the contact area between the electronic component and the ceramic packaging layer. The structure defect area of the ceramic packaging layer is also determined based on the relative position relationship information of the contact defect area and the outer surface defect area of the ceramic packaging layer, so as to globally position and identify the structure defect of the ceramic packaging layer of the electronic component, and improve the accuracy of the good product identification of the electronic component.

[0004] The present application is achieved by the following technical solutions:

[0005] A ceramic packaging basic surface defect detection method based on image recognition, comprising:

[0006] performing first visual collection on a contact area surface between the electronic component and the ceramic packaging layer to obtain a contact area surface image; performing recognition on the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer; performing analysis on the contact space state data to determine a contact abnormal area between the electronic component and the ceramic packaging layer;

[0007] performing picture extraction processing on the contact area surface image based on position information of the contact abnormal area to obtain a corresponding image sub-picture; performing recognition on the image sub-picture to obtain corresponding contact interface topography data; performing analysis on the contact interface topography data to determine a contact defect area between the electronic component and the ceramic packaging layer.

[0008] performing second visual collection on an outer surface of the ceramic packaging layer to obtain a ceramic layer outer surface image; performing recognition on the ceramic layer outer surface image to obtain an outer surface defect area of the ceramic packaging layer; determining a structural defect area of the ceramic packaging layer based on relative position relationship information of the contact defect area and the outer surface defect area.

[0009] Optionally, performing first visual collection on a contact area surface between the electronic component and the ceramic packaging layer to obtain a contact area surface image; performing recognition on the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer; performing analysis on the contact space state data to determine a contact abnormal area between the electronic component and the ceramic packaging layer, comprising:

[0010] performing scanning visual collection on a contact area surface between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining respective pixel texture distribution position data of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining respective edge contour information of the electronic component and the ceramic packaging layer on the contact area surface based on the pixel texture distribution position data; and determining contact space gap size variation data between the electronic component and the ceramic packaging layer based on the edge contour information.

[0011] performing analysis on the contact space gap size variation data, if gap sizes of a part of regions of the contact space are all greater than a preset size threshold, then determining the corresponding part of regions as the contact abnormal area between the electronic component and the ceramic packaging layer.

[0012] Optionally, based on the position information of the abnormal contact area, performing image extraction processing on the surface image of the contact area to obtain a corresponding image sub-image; identifying the image sub-image to obtain corresponding contact interface topography data; and analyzing the contact interface topography data to determine the contact defect area between the electronic component and the ceramic packaging layer, including:

[0013] Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture;

[0014] Pixel contour recognition is performed on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; based on the pixel contour distribution feature information, a corresponding contact interface morphology roughness is obtained; the contact interface morphology roughness is compared with a preset roughness threshold; if the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined as a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

[0015] Optionally, scanning and visually capturing the contact area surface between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface includes:

[0016] Step S1, using the following formula (1), according to the total area of ​​the contact area between the electronic component and the ceramic packaging layer and the maximum area of ​​a single scan for scanning visual acquisition, determine the single scan shooting area for scanning visual acquisition,

[0017]

[0018] In the above formula (1), ΔS represents the single scanning shooting area for scanning visual acquisition; S0 represents the maximum single scanning area for scanning visual acquisition; S represents the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer;

[0019] Step S2, using the following formula (2), according to the single scanning shooting area and the total area of ​​the contact area between the electronic component and the ceramic packaging layer, determine the single scanning shooting time for scanning visual acquisition.

[0020]

[0021] In the above formula (2), T represents the duration of a single scan for scanning visual acquisition; T maxIndicates the maximum duration of a single scan for scanning visual acquisition; T min Indicates the minimum duration of a single scan for scanning visual acquisition;

[0022] Step S3, using the following formula (3), according to the single scanning shooting area and the single scanning shooting duration, controls the overlapping shooting area each time when performing scanning visual acquisition,

[0023]

[0024] In the above formula (3), S c Indicates the overlapping shooting area each time during scanning visual acquisition.

[0025] Optionally, performing a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an outer surface image of the ceramic layer; identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer; and determining a structural defect area of ​​the ceramic packaging layer based on relative positional relationship information between the contact defect area and the outer surface defect area, including:

[0026] Performing binocular vision acquisition on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; generating a three-dimensional image of the outer surface of the ceramic layer based on binocular parallax of the binocular outer surface image; recognizing the three-dimensional image of the outer surface of the ceramic layer to obtain outer surface crack distribution information of the ceramic packaging layer; determining the crack distribution density of the outer surface of the ceramic packaging layer based on the outer surface crack distribution information, thereby identifying the outer surface defect area of ​​the ceramic packaging layer;

[0027] The position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer are compared to determine the closest relative distance between the contact defect area and the outer surface defect area; if the closest relative distance is less than a preset distance threshold, the area occupied by the contact defect area and the outer surface defect area is determined as the structural defect area of ​​the ceramic packaging layer; otherwise, the contact defect area and the outer surface defect area are not determined as the structural defect area of ​​the ceramic packaging layer.

[0028] A basic surface defect detection system for ceramic packages based on image recognition, comprising:

[0029] a first visual acquisition and recognition module, configured to perform a first visual acquisition of the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a contact area surface image; and recognize the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer;

[0030] a contact abnormality region determining module, configured to analyze the contact space state data and determine the contact abnormality region between the electronic component and the ceramic packaging layer;

[0031] A contact interface topography determination module is configured to perform image extraction processing on the surface image of the contact area based on the position information of the contact abnormality area to obtain a corresponding image sub-image; and to identify the image sub-image to obtain corresponding contact interface topography data;

[0032] a contact defect area determination module, configured to analyze the contact interface topography data to determine a contact defect area between the electronic component and the ceramic packaging layer;

[0033] A second visual acquisition and recognition module is used to perform a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; and recognize the image of the outer surface of the ceramic layer to obtain a defect area on the outer surface of the ceramic packaging layer;

[0034] The structural defect area determination module is used to determine the structural defect area of ​​the ceramic packaging layer based on the relative position relationship information between the contact defect area and the outer surface defect area.

[0035] Optionally, the first visual acquisition and recognition module is configured to perform a first visual acquisition of the contact area surface between the electronic component and the ceramic packaging layer to obtain a contact area surface image; and recognize the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer, including:

[0036] Performing scanning visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining pixel texture distribution position data corresponding to each of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining edge contour information of each of the electronic component and the ceramic packaging layer on the surface of the contact area based on the pixel texture distribution position data; and determining contact space gap size change data between the electronic component and the ceramic packaging layer based on the edge contour information;

[0037] The abnormal contact area determination module is used to analyze the contact space state data to determine the abnormal contact area between the electronic component and the ceramic packaging layer, including:

[0038] The contact space gap size change data is analyzed. If the gap size of a certain area of ​​the contact space is greater than a preset size threshold, the corresponding area is determined as an abnormal contact area between the electronic component and the ceramic packaging layer.

[0039] Optionally, the contact interface topography determination module is configured to perform image extraction processing on the surface image of the contact area based on the position information of the contact abnormality area to obtain a corresponding image sub-image; and identify the image sub-image to obtain corresponding contact interface topography data, including:

[0040] Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture;

[0041] Performing pixel contour recognition on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; and obtaining corresponding contact interface topography roughness based on the pixel contour distribution feature information;

[0042] The contact defect area determination module is used to analyze the contact interface topography data to determine the contact defect area between the electronic component and the ceramic packaging layer, including:

[0043] The contact interface morphology roughness is compared with a preset roughness threshold. If the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined as a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

[0044] Optionally, the second visual acquisition and recognition module is configured to perform a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; and recognize the image of the outer surface of the ceramic layer to obtain a defective area on the outer surface of the ceramic packaging layer, including:

[0045] Performing binocular vision acquisition on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; generating a three-dimensional image of the outer surface of the ceramic layer based on binocular parallax of the binocular outer surface image; recognizing the three-dimensional image of the outer surface of the ceramic layer to obtain outer surface crack distribution information of the ceramic packaging layer; determining the crack distribution density of the outer surface of the ceramic packaging layer based on the outer surface crack distribution information, thereby identifying the outer surface defect area of ​​the ceramic packaging layer;

[0046] The structural defect area determination module is configured to determine the structural defect area of ​​the ceramic packaging layer based on the relative positional relationship information between the contact defect area and the outer surface defect area, including:

[0047] The position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer are compared to determine the closest relative distance between the contact defect area and the outer surface defect area; if the closest relative distance is less than a preset distance threshold, the area occupied by the contact defect area and the outer surface defect area is determined as the structural defect area of ​​the ceramic packaging layer; otherwise, the contact defect area and the outer surface defect area are not determined as the structural defect area of ​​the ceramic packaging layer.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The image recognition-based ceramic packaging basic surface defect detection method and system provided in the present application performs a first visual acquisition and recognition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain the contact space state data between the two, and performs a global quantitative recognition of the contact tightness between the two, so as to determine the abnormal contact area between the two; then, image picture extraction and recognition are performed on the abnormal contact area, and the contact interface morphology data of the two are analyzed in more detail, and the contact change between the two is characterized to determine the contact defect area between the two, thereby locating the area where the electronic component and the ceramic packaging layer have loose contact; and second visual acquisition and recognition are performed on the outer surface of the ceramic packaging layer to obtain the outer surface defect area of ​​the ceramic packaging layer, and based on the relative position relationship information of the contact defect area and the outer surface defect area, the structural defect area of ​​the ceramic packaging layer is determined, thereby globally locating and identifying the structural defects of the ceramic packaging layer of the electronic component, thereby improving the accuracy of identifying good electronic components. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. Among them:

[0051] Figure 1 A schematic flow chart of a method for detecting basic surface defects of ceramic packages based on image recognition provided by the present invention.

[0052] Figure 2 This is a structural schematic diagram of a ceramic package basic surface defect detection system based on image recognition provided by the present invention. DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below in conjunction with the accompanying drawings. It will be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some, rather than all, structures related to the present application are shown in the accompanying drawings. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0054] As used herein, the terms "comprise," "comprising," and "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0055] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0056] See also Figure 1 As shown, an embodiment of the present application provides a method for detecting basic surface defects of ceramic packages based on image recognition. The method for detecting basic surface defects of ceramic packages based on image recognition includes:

[0057] Performing a first visual acquisition of the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a contact area surface image; recognizing the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer; and analyzing the contact space state data to determine an abnormal contact area between the electronic component and the ceramic packaging layer;

[0058] Based on the position information of the contact abnormality area, performing image extraction processing on the surface image of the contact area to obtain a corresponding image sub-image; identifying the image sub-image to obtain corresponding contact interface morphology data; analyzing the contact interface morphology data to determine the contact defect area between the electronic component and the ceramic packaging layer;

[0059] A second visual acquisition is performed on the outer surface of the ceramic packaging layer to obtain an outer surface image of the ceramic layer; the outer surface image of the ceramic layer is identified to obtain an outer surface defect area of ​​the ceramic packaging layer; and based on the relative positional relationship information between the contact defect area and the outer surface defect area, the structural defect area of ​​the ceramic packaging layer is determined.

[0060] The beneficial effects of the above embodiments are as follows: the basic surface defect detection method for ceramic packaging based on image recognition performs a first visual acquisition and recognition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain the contact space state data between the two, and performs a global quantitative recognition of the contact tightness between the two, thereby determining the abnormal contact area between the two; then, image extraction and recognition are performed on the abnormal contact area, and the contact interface morphology data of the two are analyzed in more detail, and the contact change between the two is characterized to determine the contact defect area between the two, thereby locating the area where the electronic component and the ceramic packaging layer have loose contact; and a second visual acquisition and recognition is performed on the outer surface of the ceramic packaging layer to obtain the outer surface defect area of ​​the ceramic packaging layer, and based on the relative position relationship information of the contact defect area and the outer surface defect area, the structural defect area of ​​the ceramic packaging layer is determined, thereby globally locating and identifying the structural defects of the ceramic packaging layer of the electronic component, thereby improving the accuracy of identifying good electronic components.

[0061] In another embodiment, performing a first visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a contact area surface image; recognizing the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer; and analyzing the contact space state data to determine an abnormal contact area between the electronic component and the ceramic packaging layer, including:

[0062] Performing a scanning visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a global image of the surface of the contact area; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining pixel texture distribution position data corresponding to each of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining edge contour information of each of the electronic component and the ceramic packaging layer on the surface of the contact area based on the pixel texture distribution position data; and determining contact space gap size change data between the electronic component and the ceramic packaging layer based on the edge contour information;

[0063] The data of the change in the gap size of the contact space is analyzed. If the gap size of a certain area of ​​the contact space is greater than a preset size threshold, the corresponding area is determined as an abnormal contact area between the electronic component and the ceramic packaging layer.

[0064] The beneficial effects of the above embodiments are that the ceramic packaging layer is combined with the electronic component through a high-temperature sintering process, and the degree of bonding contact between the ceramic packaging layer and the electronic component depends on the parameters of the high-temperature sintering process. In order to globally identify the degree of contact between the ceramic packaging layer and the electronic component, the surface of the contact area between the electronic component and the ceramic packaging layer (that is, the cross-sectional surface of the contact area between the electronic component and the ceramic packaging layer) is scanned and visually captured to obtain a global image of the surface of the contact area. The ceramic packaging layer and the electronic component have corresponding materials that are different, so the pixel textures of the ceramic packaging layer and the electronic component on the global image are also different. For this reason, pixel texture recognition is performed on the global image to obtain pixel texture distribution data of the entire picture area of ​​the global image. The pixel texture distribution data is then differentiated to determine the pixel texture distribution position data corresponding to the electronic component and the ceramic packaging layer. In this way, the edge contour information of the electronic component and the ceramic packaging layer on the surface of the contact area is accurately calculated, and the edge contour lines corresponding to the two on the surface of the contact area are effectively distinguished. Then, based on the edge contour lines, the contact space gap size change data between the electronic component and the ceramic packaging layer is directly determined (that is, the width of the gap space formed between the electronic component and the ceramic packaging layer in the entire length extension dimension of the contact area surface). The data on the change in the gap size of the contact space is then analyzed. If the gap size of a certain area of ​​the contact space is greater than the preset size threshold, it indicates that the gap between the electronic component and the ceramic packaging layer in this area is large (that is, the contact is not sufficiently close). At this time, the corresponding area will be determined as the abnormal contact area between the electronic component and the ceramic packaging layer, which is convenient for narrowing the scope of investigation of the contact defect position between the electronic component and the ceramic packaging layer.

[0065] In another embodiment, scanning and visually capturing the contact area surface between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface includes:

[0066] Step S1, using the following formula (1), according to the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer and the maximum area of ​​a single scan for scanning visual acquisition, determine the single scan shooting area for scanning visual acquisition,

[0067]

[0068] In the above formula (1), ΔS represents the single scanning area for scanning visual acquisition; S0 represents the maximum single scanning area for scanning visual acquisition; S represents the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer;

[0069] Step S2, using the following formula (2), according to the single scanning shooting area and the total area of ​​the contact area between the electronic component and the ceramic packaging layer, determine the single scanning shooting time for scanning visual acquisition.

[0070]

[0071] In the above formula (2), T represents the duration of a single scan for scanning visual acquisition; T max Indicates the maximum duration of a single scan for scanning visual acquisition; T min Indicates the minimum duration of a single scan for scanning visual acquisition;

[0072] Step S3, using the following formula (3), according to the single scanning shooting area and the single scanning shooting duration, controls the overlapping shooting area each time when performing scanning visual acquisition,

[0073]

[0074] In the above formula (3), S c Indicates the overlapping shooting area each time during scanning visual acquisition.

[0075] The beneficial effects of the above embodiment are as follows: using the above formula (1), according to the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer and the maximum area of ​​a single scan for scanning visual acquisition, the single scan shooting area for scanning visual acquisition is determined, thereby increasing the single scan shooting area when the total area is large to improve the efficiency of the shooting scan, and reducing the single scan shooting area when the total area is small to ensure the clarity of the shooting; then using the above formula (2), according to the single scan shooting area and the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer, the single scan shooting time for scanning visual acquisition is determined, thereby speeding up the scanning shooting when shooting a small area and improving the system efficiency; then using the above formula (3), according to the single scan shooting area and the single scan shooting time, the overlapping shooting area each time when scanning visual acquisition is performed is controlled, thereby increasing the overlapping shooting area when the shooting scanning area is large to improve the reliability of image scanning shooting.

[0076] In another embodiment, based on the position information of the contact abnormality area, performing image extraction processing on the surface image of the contact area to obtain a corresponding image sub-image; identifying the image sub-image to obtain corresponding contact interface topography data; and analyzing the contact interface topography data to determine the contact defect area between the electronic component and the ceramic packaging layer, including:

[0077] Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture;

[0078] Pixel contour recognition is performed on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; based on the pixel contour distribution feature information, a corresponding contact interface morphology roughness is obtained; the contact interface morphology roughness is compared with a preset roughness threshold; if the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined to be a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

[0079] The beneficial effects of the above embodiment are as follows: coordinate transformation analysis is performed on the position information of the contact abnormality area to determine the boundary coordinate information of the contact abnormality area, and the boundary coordinate information is used as a reference to match the boundary mapping of the surface image of the contact area, thereby segmenting and extracting the corresponding image sub-image in the contact area surface image, thereby ensuring that the image sub-image corresponds only to the contact abnormality area, facilitating subsequent more detailed image recognition of only the contact abnormality area. Pixel contour recognition is then performed on the image sub-image to obtain pixel contour distribution feature information of the image sub-image, thereby obtaining the contact interface morphology roughness of the contact abnormality area. The greater the contact interface morphology roughness, the less tight the contact between the electronic component and the ceramic packaging layer. At this time, the contact interface morphology roughness is compared with a preset roughness threshold. If the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined to be a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined to be a contact defect area. In this way, the defect area between the electronic component and the ceramic packaging layer that is not in sufficient close contact can be identified and located.

[0080] In another embodiment, performing a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an outer surface image of the ceramic layer; identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer; and determining a structural defect area of ​​the ceramic packaging layer based on relative positional relationship information between the contact defect area and the outer surface defect area, including:

[0081] Performing binocular vision acquisition on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; generating a three-dimensional image of the outer surface of the ceramic layer based on binocular parallax of the binocular outer surface image; recognizing the three-dimensional image of the outer surface of the ceramic layer to obtain outer surface crack distribution information of the ceramic packaging layer; determining the crack distribution density of the outer surface of the ceramic packaging layer based on the outer surface crack distribution information, thereby identifying the outer surface defect area of ​​the ceramic packaging layer;

[0082] The position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer are compared to determine the closest relative distance between the contact defect area and the outer surface defect area; if the closest relative distance is less than a preset distance threshold, the area occupied by the contact defect area and the outer surface defect area is regarded as the structural defect area of ​​the ceramic packaging layer; otherwise, the contact defect area and the outer surface defect area are not determined as the structural defect area of ​​the ceramic packaging layer.

[0083] The beneficial effect of the above embodiment is that during the high-temperature sintering process between the ceramic packaging layer and the electronic component, cracks will form on the outer surface of the ceramic packaging layer. When the distribution density of the cracks increases, the ceramic packaging layer will be unable to effectively seal and isolate the electronic component. To this end, binocular vision is used to collect and analyze the outer surface of the ceramic packaging layer to obtain crack distribution information on the outer surface of the ceramic packaging layer. Based on this outer surface crack distribution information, the crack distribution density of the outer surface of the ceramic packaging layer is determined. If the crack distribution density of a certain grid area on the outer surface of the ceramic packaging layer is greater than a preset distribution density threshold, the grid area is determined as an outer surface defect area of ​​the ceramic packaging layer. Otherwise, the grid area is not determined as an outer surface defect area of ​​the ceramic packaging layer. In addition, the position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer is compared to determine the closest relative distance between the contact defect area and the outer surface defect area, and then a threshold comparison is performed on the closest relative distance. If the closest relative distance is less than the preset distance threshold, it indicates that the contact defect area and the outer surface defect area are too close and easily affect each other. At this time, the area occupied by the contact defect area and the outer surface defect area is regarded as the structural defect area of ​​the ceramic packaging layer. In this way, the packaging defect area of ​​the ceramic packaging layer for the electronic component can be accurately and comprehensively identified.

[0084] See also Figure 2 As shown, an embodiment of the present application provides a ceramic package basic surface defect detection system based on image recognition. The ceramic package basic surface defect detection system based on image recognition includes:

[0085] The first visual acquisition and recognition module is configured to acquire a first visual image of a contact area surface between the electronic component and the ceramic packaging layer, and obtain contact area surface image; and recognize the contact area surface image, and obtain contact space state data between the electronic component and the ceramic packaging layer.

[0086] The contact abnormal area determination module is configured to analyze the contact space state data, and determine a contact abnormal area between the electronic component and the ceramic packaging layer.

[0087] The contact interface topography determination module is configured to perform picture extraction processing on the contact area surface image based on position information of the contact abnormal area, and obtain a corresponding image sub-picture; and recognize the image sub-picture, and obtain corresponding contact interface topography data.

[0088] The contact defect area determination module is configured to analyze the contact interface topography data, and determine a contact defect area between the electronic component and the ceramic packaging layer.

[0089] The second visual acquisition and recognition module is configured to acquire a second visual image of an outer surface of the ceramic packaging layer, and obtain ceramic layer outer surface image; and recognize the ceramic layer outer surface image, and obtain an outer surface defect area of the ceramic packaging layer.

[0090] The structural defect area determination module is configured to determine a structural defect area of the ceramic packaging layer based on relative position relationship information of the contact defect area and the outer surface defect area.

[0091] The above-mentioned embodiment has the beneficial effect that the ceramic packaging basic surface defect detection system based on image recognition acquires and recognizes a first visual image of a contact area surface between an electronic component and a ceramic packaging layer, obtains contact space state data between the two, globally quantifies and recognizes a contact tightness between the two, determines a contact abnormal area between the two, extracts and recognizes an image picture of the contact abnormal area, more finely analyzes contact interface topography data of the two, characterizes a contact change of the two, determines a contact defect area between the two, positions a contact not tight area between the electronic component and the ceramic packaging layer, acquires and recognizes a second visual image of an outer surface of the ceramic packaging layer, obtains an outer surface defect area of the ceramic packaging layer, and determines a structural defect area of the ceramic packaging layer based on relative position relationship information of the contact defect area and the outer surface defect area, globally positions and recognizes a structural defect of the ceramic packaging layer of the electronic component, and improves good product discrimination accuracy of the electronic component.

[0092] In another embodiment, the first visual acquisition and recognition module is used to perform a first visual acquisition of the contact area surface between the electronic component and the ceramic packaging layer to obtain a contact area surface image; and recognize the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer, including:

[0093] Performing a scanning visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a global image of the surface of the contact area; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining pixel texture distribution position data corresponding to each of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining edge contour information of each of the electronic component and the ceramic packaging layer on the surface of the contact area based on the pixel texture distribution position data; and determining contact space gap size change data between the electronic component and the ceramic packaging layer based on the edge contour information;

[0094] The abnormal contact area determination module is used to analyze the contact space state data to determine the abnormal contact area between the electronic component and the ceramic packaging layer, including:

[0095] The data of the change in the gap size of the contact space is analyzed. If the gap size of a certain area of ​​the contact space is greater than a preset size threshold, the corresponding area is determined as an abnormal contact area between the electronic component and the ceramic packaging layer.

[0096] The beneficial effects of the above embodiments are that the ceramic packaging layer is combined with the electronic component through a high-temperature sintering process, and the degree of bonding contact between the ceramic packaging layer and the electronic component depends on the parameters of the high-temperature sintering process. In order to globally identify the degree of contact between the ceramic packaging layer and the electronic component, the surface of the contact area between the electronic component and the ceramic packaging layer (that is, the cross-sectional surface of the contact area between the electronic component and the ceramic packaging layer) is scanned and visually captured to obtain a global image of the surface of the contact area. The ceramic packaging layer and the electronic component have corresponding materials that are different, so the pixel textures of the ceramic packaging layer and the electronic component on the global image are also different. For this reason, pixel texture recognition is performed on the global image to obtain pixel texture distribution data of the entire picture area of ​​the global image. The pixel texture distribution data is then differentiated to determine the pixel texture distribution position data corresponding to the electronic component and the ceramic packaging layer. In this way, the edge contour information of the electronic component and the ceramic packaging layer on the surface of the contact area is accurately calculated, and the edge contour lines corresponding to the two on the surface of the contact area are effectively distinguished. Then, based on the edge contour lines, the contact space gap size change data between the electronic component and the ceramic packaging layer is directly determined (that is, the width of the gap space formed between the electronic component and the ceramic packaging layer in the entire length extension dimension of the contact area surface). The data on the change in the gap size of the contact space is then analyzed. If the gap size of a certain area of ​​the contact space is greater than the preset size threshold, it indicates that the gap between the electronic component and the ceramic packaging layer in this area is large (that is, the contact is not sufficiently close). At this time, the corresponding area will be determined as the abnormal contact area between the electronic component and the ceramic packaging layer, which is convenient for narrowing the scope of investigation of the contact defect position between the electronic component and the ceramic packaging layer.

[0097] In another embodiment, the contact interface topography determination module is configured to perform image extraction processing on the surface image of the contact area based on the position information of the contact abnormality area to obtain a corresponding image sub-image; and identify the image sub-image to obtain corresponding contact interface topography data, including:

[0098] Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture;

[0099] Performing pixel contour recognition on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; and obtaining corresponding contact interface topography roughness based on the pixel contour distribution feature information;

[0100] The contact defect area determination module is used to analyze the contact interface topography data to determine the contact defect area between the electronic component and the ceramic packaging layer, including:

[0101] The contact interface morphology roughness is compared with a preset roughness threshold. If the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined as a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

[0102] The beneficial effects of the above embodiment are as follows: coordinate transformation analysis is performed on the position information of the contact abnormality area to determine the boundary coordinate information of the contact abnormality area, and the boundary coordinate information is used as a reference to match the boundary mapping of the surface image of the contact area, thereby segmenting and extracting the corresponding image sub-image in the contact area surface image, thereby ensuring that the image sub-image corresponds only to the contact abnormality area, facilitating subsequent more detailed image recognition of only the contact abnormality area. Pixel contour recognition is then performed on the image sub-image to obtain pixel contour distribution feature information of the image sub-image, thereby obtaining the contact interface morphology roughness of the contact abnormality area. The greater the contact interface morphology roughness, the less tight the contact between the electronic component and the ceramic packaging layer. At this time, the contact interface morphology roughness is compared with a preset roughness threshold. If the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined to be a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined to be a contact defect area. In this way, the defect area between the electronic component and the ceramic packaging layer that is not in sufficient close contact can be identified and located.

[0103] In another embodiment, the second visual acquisition and recognition module is used to perform a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; and recognize the image of the outer surface of the ceramic layer to obtain a defect area on the outer surface of the ceramic packaging layer, including:

[0104] Performing binocular vision acquisition on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; generating a three-dimensional image of the outer surface of the ceramic layer based on binocular parallax of the binocular outer surface image; recognizing the three-dimensional image of the outer surface of the ceramic layer to obtain outer surface crack distribution information of the ceramic packaging layer; determining the crack distribution density of the outer surface of the ceramic packaging layer based on the outer surface crack distribution information, thereby identifying the outer surface defect area of ​​the ceramic packaging layer;

[0105] The structural defect area determination module is used to determine the structural defect area of ​​the ceramic packaging layer based on the relative position relationship information between the contact defect area and the outer surface defect area, including:

[0106] The position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer are compared to determine the closest relative distance between the contact defect area and the outer surface defect area; if the closest relative distance is less than a preset distance threshold, the area occupied by the contact defect area and the outer surface defect area is regarded as the structural defect area of ​​the ceramic packaging layer; otherwise, the contact defect area and the outer surface defect area are not determined as the structural defect area of ​​the ceramic packaging layer.

[0107] The beneficial effect of the above embodiment is that during the high-temperature sintering process between the ceramic packaging layer and the electronic component, cracks will form on the outer surface of the ceramic packaging layer. When the distribution density of the cracks increases, the ceramic packaging layer will be unable to effectively seal and isolate the electronic component. To this end, binocular vision is used to collect and analyze the outer surface of the ceramic packaging layer to obtain crack distribution information on the outer surface of the ceramic packaging layer. Based on this outer surface crack distribution information, the crack distribution density of the outer surface of the ceramic packaging layer is determined. If the crack distribution density of a certain grid area on the outer surface of the ceramic packaging layer is greater than a preset distribution density threshold, the grid area is determined as an outer surface defect area of ​​the ceramic packaging layer. Otherwise, the grid area is not determined as an outer surface defect area of ​​the ceramic packaging layer. In addition, the position information of the contact defect area and the outer surface defect area in the three-dimensional space corresponding to the ceramic packaging layer is compared to determine the closest relative distance between the contact defect area and the outer surface defect area, and then a threshold comparison is performed on the closest relative distance. If the closest relative distance is less than the preset distance threshold, it indicates that the contact defect area and the outer surface defect area are too close and easily affect each other. At this time, the area occupied by the contact defect area and the outer surface defect area is regarded as the structural defect area of ​​the ceramic packaging layer. In this way, the packaging defect area of ​​the ceramic packaging layer for the electronic component can be accurately and comprehensively identified.

[0108] In general, the basic surface defect detection method and system for ceramic packaging based on image recognition performs a first visual acquisition and recognition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain the contact space state data between the two, and performs a global quantitative recognition of the contact tightness between the two, so as to determine the abnormal contact area between the two; then, image extraction and recognition are performed on the abnormal contact area, and the contact interface morphology data of the two are analyzed in more detail, and the contact changes between the two are characterized, so as to determine the contact defect area between the two, and realize the positioning of the area where the electronic component and the ceramic packaging layer have loose contact; and a second visual acquisition and recognition is performed on the outer surface of the ceramic packaging layer to obtain the outer surface defect area of ​​the ceramic packaging layer, and based on the relative position relationship information of the contact defect area and the outer surface defect area, the structural defect area of ​​the ceramic packaging layer is determined, thereby globally locating and identifying the structural defects of the ceramic packaging layer of the electronic component, thereby improving the accuracy of identifying good electronic components.

[0109] The above is only a specific embodiment of the present invention, and any other improvements made based on the concept of the present invention are considered to be within the scope of protection of the present invention.

Claims

1. A method for detecting basic surface defects of ceramic packages based on image recognition, characterized in that: include: Performing a first visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain an image of the surface of the contact area; Identifying the surface image of the contact area to obtain contact space state data between the electronic component and the ceramic packaging layer; Analyzing the contact space state data to determine an abnormal contact area between the electronic component and the ceramic packaging layer; Based on the position information of the abnormal contact area, performing image extraction processing on the surface image of the contact area to obtain a corresponding image sub-image; Identifying the image sub-frame to obtain corresponding contact interface topography data; analyzing the contact interface topography data to determine a contact defect area between the electronic component and the ceramic packaging layer; Performing a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; Identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer; determining a structural defect region of the ceramic packaging layer based on relative positional relationship information between the contact defect region and the outer surface defect region; Determining a structural defect area of ​​the ceramic packaging layer based on relative positional relationship information between the contact defect area and the outer surface defect area includes: comparing the position information of the contact defect region and the outer surface defect region in the three-dimensional space corresponding to the ceramic packaging layer to determine the closest relative distance between the contact defect region and the outer surface defect region; if the closest relative distance is less than a preset distance threshold, determining the area occupied by the contact defect region and the outer surface defect region as the structural defect region of the ceramic packaging layer; Otherwise, the contact defect region and the outer surface defect region are not determined as structural defect regions of the ceramic packaging layer.

2. The method for detecting basic surface defects of ceramic packages based on image recognition according to claim 1, characterized in that: Performing a first visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a contact area surface image; and recognizing the contact area surface image to obtain contact space state data between the electronic component and the ceramic packaging layer. Analyzing the contact space state data to determine an abnormal contact area between the electronic component and the ceramic packaging layer includes: Scanning and visually capturing the contact area surface between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining pixel texture distribution position data corresponding to each of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining edge contour information of each of the electronic component and the ceramic packaging layer on the surface of the contact area based on the pixel texture distribution position data; and determining contact space gap size change data between the electronic component and the ceramic packaging layer based on the edge contour information; The contact space gap size change data is analyzed. If the gap size of a certain area of ​​the contact space is greater than a preset size threshold, the corresponding area is determined as an abnormal contact area between the electronic component and the ceramic packaging layer.

3. The method for detecting basic surface defects of ceramic packages based on image recognition according to claim 2, characterized in that: Scan the contact area surface between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface, including: Step S1, using the following formula (1), according to the total surface area of ​​the contact area between the electronic component and the ceramic packaging layer and the maximum area of ​​a single scan for scanning visual acquisition, determine the single scan shooting area for scanning visual acquisition, （1） In the above formula (1), Indicates the single scan area for scanning visual acquisition; Indicates the maximum area of ​​a single scan for scanning visual acquisition; represents the total surface area of ​​the contact region between the electronic component and the ceramic packaging layer; Step S2, using the following formula (2), according to the single scanning shooting area and the total area of ​​the contact area between the electronic component and the ceramic packaging layer, determine the single scanning shooting time for scanning visual acquisition. （2） In the above formula (2), Indicates the duration of a single scan for scanning visual acquisition; Indicates the maximum duration of a single scan for scanning visual acquisition; Indicates the minimum duration of a single scan for scanning visual acquisition; Step S3, using the following formula (3), according to the single scanning shooting area and the single scanning shooting time, controls the overlapping shooting area each time when performing scanning visual acquisition, （3） In the above formula (3), Indicates the overlapping shooting area each time during scanning visual acquisition.

4. The method for detecting basic surface defects of ceramic packages based on image recognition according to claim 1, wherein: Based on the position information of the abnormal contact area, performing image extraction processing on the surface image of the contact area to obtain a corresponding image sub-image; Identifying the image sub-picture to obtain corresponding contact interface topography data; Analyzing the contact interface topography data to determine a contact defect area between the electronic component and the ceramic packaging layer includes: Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture; Pixel contour recognition is performed on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; based on the pixel contour distribution feature information, a corresponding contact interface morphology roughness is obtained; the contact interface morphology roughness is compared with a preset roughness threshold; if the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined as a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

5. The method for detecting basic surface defects of ceramic packages based on image recognition according to claim 1, wherein: Performing a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; Identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer; Determining a structural defect area of ​​the ceramic packaging layer based on relative positional relationship information between the contact defect area and the outer surface defect area includes: Binocular vision is performed on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; based on the binocular parallax of the binocular outer surface image, a three-dimensional image of the outer surface of the ceramic layer is generated; the three-dimensional image of the outer surface of the ceramic layer is recognized to obtain outer surface crack distribution information of the ceramic packaging layer; based on the outer surface crack distribution information, the crack distribution density of the outer surface of the ceramic packaging layer is determined to identify the outer surface defect area of ​​the ceramic packaging layer.

6. A ceramic package basic surface defect detection system based on image recognition, characterized in that: include: A first visual acquisition and recognition module is used to perform a first visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain an image of the contact area surface; Identifying the surface image of the contact area to obtain contact space state data between the electronic component and the ceramic packaging layer; a contact abnormality region determining module, configured to analyze the contact space state data and determine the contact abnormality region between the electronic component and the ceramic packaging layer; A contact interface topography determination module is configured to perform image extraction processing on the surface image of the contact area based on the position information of the contact abnormality area to obtain a corresponding image sub-image; Identifying the image sub-picture to obtain corresponding contact interface topography data; a contact defect area determination module, configured to analyze the contact interface topography data to determine a contact defect area between the electronic component and the ceramic packaging layer; A second visual acquisition and recognition module is used to perform a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; Identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer; a structural defect region determining module, configured to determine the structural defect region of the ceramic packaging layer based on information on a relative positional relationship between the contact defect region and the outer surface defect region; The structural defect area determination module is configured to determine the structural defect area of ​​the ceramic packaging layer based on the relative positional relationship information between the contact defect area and the outer surface defect area, including: comparing the position information of the contact defect region and the outer surface defect region in the three-dimensional space corresponding to the ceramic packaging layer to determine the closest relative distance between the contact defect region and the outer surface defect region; if the closest relative distance is less than a preset distance threshold, determining the area occupied by the contact defect region and the outer surface defect region as the structural defect region of the ceramic packaging layer; Otherwise, the contact defect region and the outer surface defect region are not determined as structural defect regions of the ceramic packaging layer.

7. The ceramic package basic surface defect detection system based on image recognition according to claim 6, characterized in that: The first visual acquisition and recognition module is used to perform a first visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain an image of the contact area surface; Identifying the surface image of the contact area to obtain contact space state data between the electronic component and the ceramic packaging layer includes: Performing scanning visual acquisition on the surface of the contact area between the electronic component and the ceramic packaging layer to obtain a global image of the contact area surface; performing pixel texture recognition on the global image to obtain pixel texture distribution data on the global image, and determining pixel texture distribution position data corresponding to each of the electronic component and the ceramic packaging layer based on the pixel texture distribution data; determining edge contour information of each of the electronic component and the ceramic packaging layer on the surface of the contact area based on the pixel texture distribution position data; and determining contact space gap size change data between the electronic component and the ceramic packaging layer based on the edge contour information; The abnormal contact area determination module is used to analyze the contact space state data to determine the abnormal contact area between the electronic component and the ceramic packaging layer, including: The contact space gap size change data is analyzed. If the gap size of a certain area of ​​the contact space is greater than a preset size threshold, the corresponding area is determined as an abnormal contact area between the electronic component and the ceramic packaging layer.

8. The ceramic package basic surface defect detection system based on image recognition according to claim 6, characterized in that: The contact interface topography determination module is used to perform image extraction processing on the surface image of the contact area based on the position information of the contact abnormal area to obtain a corresponding image sub-image; Identify the image sub-frame to obtain corresponding contact interface topography data, including: Based on the position information of the abnormal contact area, the boundary coordinate information of the abnormal contact area is determined, and the surface image of the contact area is matched with the boundary mapping, thereby segmenting and extracting the surface image of the contact area to obtain a corresponding image sub-picture; Performing pixel contour recognition on the image sub-picture to obtain pixel contour distribution feature information of the image sub-picture; and obtaining corresponding contact interface topography roughness based on the pixel contour distribution feature information; The contact defect area determination module is used to analyze the contact interface topography data to determine the contact defect area between the electronic component and the ceramic packaging layer, including: The contact interface morphology roughness is compared with a preset roughness threshold. If the contact interface morphology roughness is greater than the preset roughness threshold, the contact abnormality area is determined as a contact defect area between the electronic component and the ceramic packaging layer; otherwise, the contact abnormality is not determined as a contact defect area.

9. The ceramic package basic surface defect detection system based on image recognition according to claim 6, characterized in that: The second visual acquisition and recognition module is used to perform a second visual acquisition on the outer surface of the ceramic packaging layer to obtain an image of the outer surface of the ceramic layer; Identifying the outer surface image of the ceramic layer to obtain an outer surface defect area of ​​the ceramic packaging layer includes: Binocular vision is performed on the outer surface of the ceramic packaging layer to obtain a binocular outer surface image; based on the binocular parallax of the binocular outer surface image, a three-dimensional image of the outer surface of the ceramic layer is generated; the three-dimensional image of the outer surface of the ceramic layer is recognized to obtain outer surface crack distribution information of the ceramic packaging layer; based on the outer surface crack distribution information, the crack distribution density of the outer surface of the ceramic packaging layer is determined to identify the outer surface defect area of ​​the ceramic packaging layer.

Images